Updating the Pharmacological Effects of α-Mangostin Compound and Unraveling Its Mechanism of Action: A Computational Study Review

Introduction

The increased progress in scientific knowledge, particularly in the realm of health, opens up extensive opportunities for delving into novel drug development. The prevalence of diverse physiological and pathological disorders in the body that cannot be effectively addressed, enlarges the demand for the discovery of more effective and safe medications. The strategy in the pursuit of drug discovery has now shifted towards the utilization of secondary metabolite compounds from natural sources.^1–3 Currently, over 50% of available marketed drugs originate from compounds found in natural materials. This approach not only enables researchers to explore the biodiversity of our planet, but also offers a promising alternative for the development of more effective medicines.^4–7 With the continued growth of research in this field, it is expected that more innovative drugs will be discovered from these diverse natural sources, furthering the enhancement of human health.

Amidst numerous potential secondary metabolite compounds, α-mangostin stands out with remarkable promise.^8–10 α-mangostin is a xanthone derivative compound commonly found in the pericarp of the mangosteen fruit (Garcinia mangostana L).^11,12 This plant is abundant and readily available, particularly in the Asian regions including Thailand, Indonesia, Malaysia, Vietnam, and India.^13,14 Not only does it have high potential as a natural resource, but α-mangostin also possesses versatile and promising pharmacological properties. Its primary efficacy is attributed to its robust antioxidant activity, which holds implications for improving various physiological conditions related to oxidative stress.^15–19 In various in vitro and in vivo studies, α-mangostin has demonstrated significant potential in the treatment of various types of cancer.²⁰ One of the most noteworthy findings is its efficacy in breast cancer cell line testing, where α-mangostin has been shown to induce apoptosis by inhibiting ER autophagy.²¹ This finding has also been confirmed in in vivo tests using HT-29 xenograft tumor-BALB/c mice, which revealed that α-mangostin contributes to the reduction of Bcl-2 and β-catenin in tumors.²² However, ongoing exploration of the pharmacological effects of α-mangostin is essential to establish its clinical efficacy.

In the relentless pursuit of uncovering the various therapeutic potential of α-mangostin, computational studies—mainly in the field of structural bioinformatics and computational biology—emerge as a strategic cornerstone, providing numerous advantages. Harnessing the power of computer-based research in the initial phases enables the prediction of the most potent pharmacologically active compounds based on their molecular mechanisms against target receptors in related pathologies.^23–25 This approach not only expedites the research timeline, potentially reducing it from 10 to 15 years, but also yields a substantial cost reduction of nearly 50%, as it allows the early elimination of compounds with low predicted potential.^26–28 Several marketed drugs, such as captopril, dorzolamide, saquinavir, indinavir, ritonavir, and tirofiban, provide evidence of the benefits of using computational studies in drug development.^29–32 Moreover, the use of computational tools has proven effective in optimizing natural product drug development, as exemplified by the case of artemisinin. Computational studies have successfully guided research efforts to obtain artemisinin derivatives with enhanced efficacy and improved pharmacokinetic profiles, such as dihydroartemisinin and artesunate.³³ In addition, the emergence of regulations minimizing or even eliminating the use of animal testing in research opens the door to the utilization and optimization of computer-aided drug design (CADD) in drug design and discovery.^34–37

Based on the presented facts, it is crucial to understand the progress in the discovery and development of candidate drug compounds based on in silico studies. Similarly, computational studies have significantly contributed to research on α-mangostin. In the search for transthyretin inhibitors for the treatment of amyloidosis, molecular docking analysis has shown that α-mangostin possesses potential, as demonstrated by its interaction with the crystal structure stored in the Protein Data Bank (PDB), under the code 4Y9C (Figure 1).³⁸ The exploration of the pharmacological potential of α-mangostin can still be further developed with the aid of computational tools. This review aims to summarize and interpret the various properties of α-mangostin based on research that utilizes computational methods. Thus, this literature review is expected to serve as a computational guide and evidence for researchers in developing α-mangostin as a drug for advanced-phase trials until it obtains the label of a marketed drug.

Figure 1 3D molecular interaction between α-mangostin and the active residue amino acids in transthyretin, visualized using Biovia Discovery Studio based on the complex crystal structure with PDB ID 4Y9C.

Method

The compilation of this review was conducted based on a thorough examination of research articles obtained from various databases, including PubMed, Scopus, and ScienceDirect. Article searches were carried out using the keywords: “(Mangostin OR α-mangostin) AND (Docking OR “In Silico” OR Computational OR Modelling)”. The obtained articles were subsequently managed based on the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guideline, starting with the removal of duplicate articles, screening for article relevance based on titles and abstracts, screening of full-text articles, and eligibility screening based on data conformity with inclusion criteria. The inclusion criteria encompassed articles in the English language, computational studies in whole or in part of the research, as well as providing detailed information on the execution of molecular docking studies, such as the software used and binding energy outcomes (binding affinity). The workflow for the search and inclusion screening process is depicted in Figure 2. Based on the information obtained from each included article, the pharmacological activities of alpha-mangostin and its molecular mechanisms, as revealed by computational studies, are discussed in detail.

Figure 2 Flowchart of literature screening based on PRISMA guideline. Notes: PRISMA figure adapted from Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. Creative Commons.39

α-Mangostin as Potential Drug Candidate

α-Mangostin is a xanthone derivative compound and one of the significant phytochemical constituents found in the tropical fruit Garcinia mangostana L.¹¹ It has also been found in other Garcinia species, including Garcinia dulcis, Garcinia staudtii, Garcinia merguensis, and Garcinia cowa.^40,41 In Garcinia mangostana, α-mangostin is predominantly located in the fruit pericarp.⁴² Chemically, α-mangostin has been identified as 3,6,8-trihydroxy-2-methoxy-1,7-bis(3-methyl but-2-enyl) xanthan-9-one.^43,44 The chemical structure of α-mangostin is shown in Figure 3. This discussion will focus on the physicochemical properties, drug-likeness properties, pharmacokinetics, and toxicity profile of α-mangostin, as predicted by computational studies.

Figure 3 Chemical structure of α-mangostin. Re-drawn using Marvin JS by Chemaxon from PubChem database.

Physicochemical and Drug-Likeness Properties

Table 1 shows the physicochemical properties of α-mangostin as determined through computational studies. Several physicochemical characteristics play a crucial role in transforming a molecule into one with drug-like properties: 1) the molecule should possess a size conducive for systemic distribution, 2) be hydrophilic enough for solubility in the bloodstream, 3) exhibit sufficient lipophilicity to traverse lipid barriers within the body, and 4) contain an adequate number of polar groups for receptor binding, but not an excessive amount that would prompt rapid elimination through urinary excretion, thereby hindering its therapeutic effect.⁴⁵ Evaluating drug-like properties, in accordance with Lipinski’s Rule of Five, entails an assessment of molecular weight, the count of hydrogen bond donors and acceptors, and the log P value. As per Lipinski’s Rule of Five, any compound deemed drug-like should possess a molecular weight under 500 Da, a partition coefficient (log P) value below 5, fewer than 5 hydrogen bond donors, and fewer than 10 hydrogen bond acceptors.⁴⁶

Table 1 Physicochemical and Drug-Likeness Properties and Pharmacokinetic and Toxicity Profile of α-Mangostin

Adherence to Lipinski’s Rule of Five provides a pragmatic framework for identifying potential drug-like compounds, as those satisfying these criteria are likely to exhibit improved folding, polarity, and molecular size.⁴⁹ It is crucial to note that the rule should not be regarded as an inflexible standalone screening criterion in isolation but rather as a dependable indicator of the potential for successful oral exposure to enhanced chemical compounds.⁵⁰ In summary, α-mangostin appears to align with all of Lipinski’s criteria for drug-likeness. These findings collectively underscore its potential as a promising therapeutic agent for a range of disorders.

Pharmacokinetic Profile

The pharmacokinetic properties of α-mangostin were predicted using the pre-ADMET and pkSCM webserver,^46–48 and the results are summarized in Table 1. These properties encompass aspects of absorption, distribution, metabolism, and excretion. In terms of absorption, α-mangostin is anticipated to exhibit high absorption levels, as indicated by the percentage of human intestinal absorption (%HIA) prediction (falling within the range of 70–100%)⁵¹ and moderate absorption according to the Caco2 cell permeability model (within the range of 4–70%).⁵² α-mangostin is associated with a relatively low volume of distribution, characterized by a score below −0.15.^46,48 Furthermore, its distribution process involves a high degree of plasma protein binding (>90%).⁴⁷ Additionally, α-mangostin is projected to breach the blood-brain barrier (BBB) with a penetration value exceeding 2.0, affirming its potential to reach the central nervous system.⁵³ This supports the notion that α-mangostin has the potential to exert positive effects on pathologies within the central nervous system.

Furthermore, α-mangostin is anticipated to undergo metabolism by CYP450 enzymes, as it is not deemed eligible as a CYP2D6 substrate.⁴⁶ Additionally, the incompatibility profile of α-mangostin as a CYP2D6 inhibitor suggests that it is not likely to negatively interact with drugs metabolized by CYP2D6 that might be co-administered.⁴⁸ Importantly, the molecule is expected to be excreted moderately, resulting in a circulating concentration in the circulation that is neither too low nor excessively high.⁴⁸ In summary, the pharmacokinetic profile of α-mangostin indicates its potential suitability as a therapeutic agent, whether administered orally or through other routes.

Toxicity Profile

The toxicity profile of α-mangostin was predicted using the same software as employed for the pharmacokinetic profile prediction. Overall, α-mangostin exhibits a favorable toxicity profile. Notably, it does not exhibit hepatotoxic, carcinogenic, or mutagenic properties. Furthermore, α-mangostin also demonstrates a relatively low predicted LD₅₀ value.^46–48 This LD₅₀ prediction aligns with findings from preclinical studies indicating that α-mangostin is non-toxic, with LD₅₀ values ranging from 500 to 5000 mg/kg body weight.⁵⁴

Molecular Docking Software Used in Investigating Pharmacological Activities of α-Mangostin

The application of computational studies in both investigating and evaluating the pharmacological effects of α-mangostin encompasses various software tools, each with diverse types and principles. Each software tool possesses its own advantages and limitations when compared to others. In this review, a variety of software tools utilized in α-mangostin research, along with their respective characteristics and interpretations, are summarized and presented in Table 2.

Table 2 Molecular Docking Software Used in Investigating α-Mangostin’s Pharmacological Activities

ArgusLab

ArgusLab is a widely used software in molecular docking studies. Initially, ArgusLab was developed for molecular modelling and molecular structure analysis. However, with the evolution of computer-based research, ArgusLab has incorporated molecular docking features. This innovation was driven by the need to address the challenges encountered in computer-aided drug discovery, particularly in the domain of protein-ligand docking.⁵⁵

One of the significant challenges faced by scientists in the application of computational studies is the cost associated with proprietary software. Addressing this financial hurdle, ArgusLab emerged as a free software solution for computational research, specifically in molecular docking. Beyond its cost-effectiveness, ArgusLab boasts advantages over other tools in terms of user-friendliness. It is highly user-friendly and can be operated even by individuals new to computational studies.⁶¹

To date, numerous studies have utilized ArgusLab in the pursuit for new drug candidates based on computational assistance. Most recently, Eawsakul et al employed this application in the exploration and optimization of astilbin, a compound found in Bauhinia strychnifolia Craib stems, as a candidate for an alpha-glucosidase inhibitor.⁶² In their investigation, the binding free energy emerged as the primary scoring function, offering a basis for comparing the activity of the tested compounds against standard compounds.⁶³ Eawsakul et al demonstrated a strong correlation between computational and in vitro studies. The findings indicate that astilbin exhibits superior alpha-glucosidase inhibitory activity (binding free energy: −7.24823 kcal/mol and IC₅₀: 22.51±0.70 μg/mL) compared to acarbose (binding free energy: −7.58662 kcal/mol and IC₅₀: 190 ± 6.97 μg/mL).⁶²

AutoDock

In parallel to ArgusLab, AutoDock stands out as a widely embraced and freely accessible software tool that has become the most popular choice for molecular docking studies. AutoDock employs the calculation of binding free energy values (in kcal/mol) as a scoring function.⁶⁴ What sets AutoDock apart from its counterparts is its distinctive edge in terms of more advanced parameterization, elevating its precision and versatility in unraveling intricate molecular interactions.⁶⁵

Before commencing docking, the binding location of the ligand to the receptor can be validated using re-docking procedures or AutoLigand.^66,67 Re-docking involves re-docking a natural compound within a complex to obtain an ideal scoring function and root-mean-square deviation (RMSD) value (considered ideal if the RMSD value is below 3.0).⁶⁶ AutoLigand is an additional feature that allows operators to instruct the software to locate the active site of the desired target receptor. Indirectly, the use of AutoLigand also enables the analysis of the active site of the target receptor.⁶⁷ However, it is worth noting that the application of AutoDock has limitations in that it can only be used for docking small molecules up to 32 rotatable bonds.⁶⁴

Recent research utilizing AutoDock was conducted by Jiang et al to investigate the molecular mechanisms of Danzhi tiaochi decoction, a traditional Chinese medicine (TCM), in the context of metabolic-associated fatty liver disease (MAFLD).⁶⁸ Since traditional medicine comprises a matrix of varying active compounds, the mechanisms of action of TCMs can be diverse, including eliminative, additive, or synergistic effects. Through the application of AutoDock, the study yielded guidance that identified five potential active compounds with optimal binding free energies to the Jun protein. In vitro and in vivo tests demonstrated a strong correlation with the results of the molecular docking study, where these five potential compounds exhibited the potential to ameliorate MAFLD via the JNK/Akt pathway. This was evident from the increased phosphorylation of c-Jun N-terminal kinase (JNK) in the MAFLD animal model.⁶⁸

AutoDock Vina

AutoDock Vina emerges as an alternative to AutoDock, offering a more user-friendly interface.⁶⁹ AutoDock Vina is designed with default system operations, simplifying the process for users and eliminating the need for complex molecular parameterization and docking configurations.⁵⁷ While this design choice places limitations on the customization of docking simulations, AutoDock Vina still provides substantial value in drug design and discovery.⁶⁴ As evidence of its efficacy, AutoDock Vina has been successfully utilized to guide the search for potential compounds within Laurus nobilis essential oil for their neuroprotective and anti-amnesic effects on memory deficits.⁷⁰

In this study, the strategic utilization of AutoDock Vina played a pivotal role in steering the discovery process, pinpointing 1.8 cineol and α-terpinyl acetate as standout compounds showcasing significant potential in inhibiting the activity of the acetylcholinesterase (AChE) enzyme. Among these, α-terpinyl acetate emerged as the most promising compound, with a binding free energy of −7.2 kcal/mol, surpassing both 1.8 cineol and donepezil, which exhibited similar binding free energies of −3.1 kcal/mol.⁷⁰

Glide

Grid-based ligand docking with energetics (Glide) represents a more advanced tool in molecular docking. As its name suggests, unlike the three previously discussed tools, Glide is predominantly used for molecular docking with a flexible docking program. Glide employs a grid-based approach to screen conformations between ligands and target proteins.⁵⁸ In general, the screening mechanism performed by Glide proceeds through several stages, beginning with site-point screening, followed by diameter screening, subset analysis, grid scoring, and refinement.^58,71 The best pose is then subjected to further analysis through grid minimization combined with Monte Carlo sampling. Finally, the output of the analysis is represented as the Glide score, a scoring function. Generally, a Glide score of −10 or lower signifies an optimal binding interaction between the ligand and its receptor.⁷² Due to its complex screening stages, the results obtained using Glide are typically more representative compared to those obtained with rigid docking programs. Nevertheless, the use of Glide is highly restricted due to its proprietary nature and the requirement for relatively high hardware specifications.⁷¹

Nonetheless, the application of Glide has proven to play a significant role in the advancement of Computer-Aided Drug Design (CADD). Most recently, Glide was utilized to confirm the activity of novel morpholino-1,3,5-triazine-pyrimidine hybrid compounds as alternatives to gliptin compounds in the management of type 2 diabetes mellitus (T2DM) by inhibiting dipeptidyl peptidase-4 (DPP-4), an ubiquitous enzyme in glucose homeostasis.⁷³ The analysis results indicated alignment between in vitro and in silico analyses. Both approaches demonstrated that one of the tested hybrid compounds exhibited superior inhibitory effects compared to Alogliptin. Through molecular docking studies, the favorable outcomes with the hybrid compounds were symbolized by their lower Glide energy values compared to the standard compounds.⁷³

Hex

In contrast to the other previously described docking tools, Hybrid Energy Exchange (HEX) is a web-based application. Its usage is relatively straightforward and user-friendly, requiring less hardware specifications.⁷⁴ HEX employs a non-3D grid-based searching system, performing searches using the Fast Fourier Transform (FFT) correlation approach.⁵⁹ Consequently, HEX offers a simpler user experience, as it does not necessitate pre-parameterization of grid coordinates. Additionally, HEX has the advantage of facilitating docking between protein-protein interactions.⁷⁵

In a study conducted by Raissa et al, HEX was effectively utilized to predict the compound most responsible for inducing apoptosis through the death receptor-5 (DR-5) induction mechanism from Azadirachta indica Juss.⁷⁶ The investigation results indicated that nimbolide exhibited superior potential compared to gedunin in inducing DR-5. This is evidenced by its lower binding energy (−247.7 vs −237.12 kcal/mol). A lower binding energy suggests a more stable interaction between the ligand and the target receptor, leading to a stronger induction effect.⁷⁶

Moe

Molecular operating environment (MOE) is a software designed for structure-based docking, sharing several similarities with AutoDock and ArgusLab in terms of its operation. However, MOE utilizes two primary parameters as scoring functions. The first, S-score, represents the docking score and can be used to predict binding affinity. The second, energy of conformation, is employed to predict the spontaneity of the reaction between the ligand and receptor in forming a stable conformation. Unlike AutoDock and ArgusLab, MOE can only be used with a registered paid license.⁶⁰

In its application, MOE has been leveraged to expedite the search for potential compounds as candidates against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). In their research, Senobari et al identified chromone-embedded peptidomimetics and furopyrimidines as potential anti-SARS-CoV-2 agents. These findings were based on the acquisition of negative S-scores, indicating strong binding affinity with the main protease and spike protein of SARS-CoV-2.⁷⁷

Various Pharmacological Activities of α-Mangostin Based on in silico Study

The search results indicate a noteworthy development in the application of computational studies to unraveling the pharmacological properties of α-mangostin. A summary of the pharmacological properties, target actions, molecular docking software, and binding affinity of α-mangostin to related target actions is presented in Table 3. Here is a detailed explanation of the pharmacological properties of α-mangostin based on the analysis of included articles.

Table 3 Pharmacological Activities and Its Mechanism of Action of α-Mangostin Based on Computational Studies

Alzheimer’s Disease

Figure 4 visually represent the role of α-mangostin in addressing Alzheimer’s disease. In the context of Alzheimer’s disease, a neurodegenerative disorder characterized by neuronal dysfunction in the brain, α-mangostin demonstrates the ability to intervene in the aggregation of β-amyloid (Aβ). α-mangostin binds strongly to Aβ with a binding affinity of −68.76 kcal/mol.^78,80 This strong binding enhances the stability of the Aβ conformation, preventing Aβ from interacting with each other.^107,108

Figure 4 Mechanism of action of α-mangostin in treating Alzheimer disease based on computational studies.

In addition to its primary mechanism in Alzheimer’s pathophysiology, α-mangostin also exhibits strong interactions with cholinesterase.⁷⁸ In Alzheimer’s conditions, cholinesterase enzyme activity leads to the degradation of acetylcholine, an essential neurotransmitter for brain function.¹⁰⁹ The interaction with the enzyme’s substrate site results in the inhibition of acetylcholine breakdown.¹¹⁰ Furthermore, through computational studies, α-mangostin is found to have a strong binding affinity for the activator site of sirtuin 1 (SIRT-1).⁷⁹ The binding strength (−9.31 kcal/mol) is significantly higher than that of resveratrol (−7.57 kcal/mol), a compound known for its ability to activate SIRT-1, and memantine (−7.49 kcal/mol), an Alzheimer’s marketed drug. Activation of SIRT-1 plays a pivotal role in Alzheimer’s condition as it possesses a remarkable capability to disrupt cross-linkages within tau proteins ensnared in neurofibrillary tangle (NFT), possibly through protein deacetylation.^111,112

Antidiabetic

In the pathology of diabetes mellitus, α-mangostin demonstrates significant potential as anti-diabetic agent through various mechanisms (Figure 5), including the inhibition of polysaccharide and disaccharide breakdown.^18,48,83 Polysaccharide breakdown is inhibited through the inhibition of the α-amylase enzyme, preventing the cleavage of α-1,4-glycosidic bonds in the polysaccharide chain.¹¹³ Disaccharide breakdown inhibition is achieved by inhibiting the α-glucosidase enzyme, which plays a role in the final breakdown of glucose in the small intestine.¹¹⁴ This finding shows that α-mangostin can inhibit the glucose breakdown process from early to late stages.

Figure 5 Mechanism of action of α-mangostin as antidiabetic agent predicted using computational studies.

Furthermore, α-mangostin’s role is seen from another etiological perspective of diabetes mellitus. Histone deacetylase 2 (HDAC2) plays a crucial role in the epigenetic aspects of diabetes mellitus. HDAC2 catalyzes the deacetylation of histone proteins on the DNA sequence, which are then transcribed and translated into glucose metabolism inhibitory factors.^115,116 α-mangostin has the potential to be an HDAC2 inhibitor candidate with a binding affinity of −7.15 kcal/mol.⁸²

On the other hand, α-mangostin also has the potential to address complications that may arise from diabetes. Excessive blood glucose levels in diabetes can lead to a shift in glucose metabolism activation to sorbitol by aldose reductase.¹¹⁷ This metabolic process involves the excessive utilization of NADPH, leading to a decrease in glutathione (GSH) and the production of reactive oxygen species (ROS), resulting in a hyperoxidative condition.¹¹⁸ α-Mangostin is recognized for its potent activity by robustly engaging with the active site of aldose reductase, with a binding affinity of −8.15 kcal/mol.⁸¹ This binding strength is comparable to the marketed drug Zopolrestat which exhibits a binding affinity of −12.59 kcal/mol.

Antioxidant

The paramount quality of α-mangostin often underscored by researchers is its noteworthy antioxidant properties. However, there is a limited number of computational studies that directly investigate α-mangostin’s role in pro-oxidant markers. Only one study conducted by Widowati et al suggests that α-mangostin may act as an antioxidant by inhibiting matrix metalloproteinase 1 (MMP-1), nuclear export protein (NEP), and prophenoloxidase (PPO3).⁸⁹ The highest binding affinity of α-mangostin is observed in its interaction with MMP-1 (−8.9 kcal/mol). NEP and PPO3 display binding affinities of −7.4 and −6.5 kcal/mol, respectively. Inhibition of MMP-1 has the secondary effect of reducing oxidative imbalance by impeding the degradation of interstitial collagen.^119,120

Breast Cancer, Oral Cancer, and Other Types of Cancer

In the case of cancer, α-mangostin has been extensively studied as an anti-breast cancer agent. However, computational studies open up the possibility of α-mangostin being used in the treatment of various other types of cancer. In the treatment of breast cancer, α-mangostin is known to actively inhibit epidermal growth factor receptor (EGFR),⁹⁰ glycogen synthase kinase 3β (GSK3β),⁹¹ insulin-like growth factor 1 receptor (IGF1R),⁹⁰ human estrogen receptor alpha (hERα),^92–94 and the β-ketoacyl part of fatty acid synthase (FAS) as depicted in Figure 6.⁹⁵ α-mangostin exhibits strong molecular interactions (binding affinity below −5.50 kcal/mol) with all of these targets.¹²¹ Although α-mangostin has a wide range of targets, the strongest predicted interaction occurs with FAS. This aligns with several findings indicating that α-mangostin is effective in reducing breast cancer malignancy through alterations in the FAS cell signaling pathway.¹²²

Figure 6 Computational studies reveal mechanism of action of α-mangostin in cancer treatment.

For oral cancer, α-mangostin has been simulated to dock with four main targets in oral cancer: calmodulin 3 (CALM3), beta arrestin 1 (ARRB1), huntingtin (HTT), and filamin A (FLNA).¹⁰⁴ α-mangostin exhibits negative binding affinity, indicating spontaneous interactions. However, it only achieves a binding affinity below −5.50 kcal/mol in the case of CALM3 (Figure 6). CALM3 binds to phosphoinositide 3-kinase α (PI3Kα), leading to the hyperactivation of the PI3Kα/Akt/mammalian target of rapamycin signalling pathway, affecting cell proliferation, differentiation, motility, and development.¹²³ The interaction between α-mangostin and CALM3 leads to downregulation of cell signalling, thereby reducing cancer cell growth progression.¹²⁴

Furthermore, in various types of cancer, overexpression of certain kinase enzymes, such as microtubule-affinity-regulating kinase 4 (MARK4) and ubiquitin-specific protease 45 (USP45), has been observed. MARK4 plays a crucial role in phosphorylating tau proteins, resulting in increased cell migration and polarity, potentially leading to metastasis.¹²⁵ Inhibition of MARK4 activity can prevent cancer cell metastasis. Additionally, inhibiting USP45 leads to the failure of cell migration by blocking the deubiquitination of SPDL1 protein and downregulating its protein expression level.¹²⁶ Both MARK4 and USP45 are highly promising as general cancer treatment targets. Interestingly, studies conducted by Khan et al⁹⁶ and Tu et al⁹⁷ demonstrate that α-mangostin strongly interacts with both of these targets. α-mangostin exhibits strong binding affinity with MARK4 and USP45, with binding affinities of −9.2 and −8.4 kcal/mol, respectively.

Chronic Kidney Disease

Indications arising from the various protective and pharmacological effects of α-mangostin suggest its capability to shield and induce recovery from organ damage, particularly in the case of chronic kidney disease (CKD). The antioxidant and anti-inflammatory effects of α-mangostin are likely the primary factors contributing to its performance in improving CKD conditions. Through an in silico study conducted by Rana et al, it is ultimately revealed that α-mangostin is predicted to play a role in ameliorating CKD conditions by altering inositol 1,4,5-trisphosphate (IP3) receptor (IP3R), kelch-like ECH-associated protein-1 (KEAP-1), c-Jun N-terminal kinase (JNK), p38 kinase, and nuclear factor kappa beta (NF-kB), as illustrated in Figure 7.⁹⁹ The study demonstrates strong binding affinities for each receptor, measuring at −3.357 kcal/mol, −5.63 kcal/mol, −6.707 kcal/mol, −4.569 kcal/mol, and −7.187 kcal/mol, respectively.

Figure 7 α-mangostin are predicted to be involved in chronic kidney disease amelioration.

Inhibition of IP3R leads to the hindrance of calcium ion (Ca²⁺) release from the endoplasmic reticulum (ER), thereby impeding apoptosis and mitochondrial damage in nephron cells.¹²⁷ Subsequently, the protective mechanism against inflammatory effects on kidney nephron cells occurs through the inhibition of the MAPK pathway (JNK and p38 kinase) and NF-kB.¹²⁸ Moreover, through its antioxidant mechanism, α-mangostin enhances the activity of endogenous antioxidant enzymes, such as sodium dismutase (SOD) and catalase (CAT). The improvement of the pro-oxidant profile is achieved indirectly due to KEAP-1 activation, which leads to the release of Nrf-2, indirectly increasing the expression of pro-oxidant enzymes.¹²⁹

Chronic Periodontitis

With its recognized anti-inflammatory properties, the prospect of utilizing α-mangostin in the treatment of chronic periodontitis emerges as a feasible avenue. However, there is limited evidence regarding the use of α-mangostin for this condition, whether in preclinical or clinical studies. Therefore, the use of computational studies is highly beneficial for predicting the efficacy and potential mechanisms of α-mangostin in the treatment of chronic periodontitis.

A study conducted by Saputra et al suggests that α-mangostin shows significant alterations with good binding affinities (below −5.50 kcal/mol) with alkaline phosphatase (ALP), osteonectin, receptor activator nuclear kappa beta (RANK) and ligand (RANKL), sclerostin, and tartrate-resistant acid phosphatase (TRAP).¹⁰⁰ Of these six targets, RANKL appears to be the most optimal target for α-mangostin in the management of chronic periodontitis, primarily due to its greater binding affinity compared to other receptors, amounting to −7.2 kcal/mol.

Infectious Disease

Based on the in silico study evidence concerning infectious diseases, α-mangostin exhibits a broad spectrum of antimicrobial effects, spanning from antibacterial, antifungal, anti-amoeba, to antiviral actions (Figure 8). As an antibacterial agent, α-mangostin is known to possess effective inhibitory activity against β-lactamase in Staphylococcus aureus bacteria.⁸⁴ The interactions involved are strong, displaying a binding affinity of −6.02 kcal/mol. Additionally, α-mangostin also demonstrates robust interactions with β-tubulin in Acanthamoeba keratitis, an amoeba species targeting the cornea, with a binding affinity of −11.22 kcal/mol.⁸⁵ Inhibition of β-tubulin’s performance obstructs the movement of essential organelles critical for the growth and replication of eukaryotic cells such as amoebas.¹³⁰

Figure 8 α-mangostin as antimicrobial agent in treating infectious diseases based on computational studies.

Beyond single-celled organisms, α-mangostin is predicted to act as an antifungal agent by inhibiting glucosamine-6-phosphate (GlcN-6-P) synthase performance.⁸⁶ Inhibition of GlcN-6-P synthase prevents fungal cells from producing N-glycosylated mannoproteins crucial for maintaining fungal cell wall integrity.^131,132 Although the interaction strength between α-mangostin and GlcN-6-P synthase is not as robust as that exhibited by terbinafine, a marketed antifungal drug (−7.04 vs −8.58 kcal/mol), it is suggested that α-mangostin may serve as an alternative antifungal treatment, especially for patients resistant to terbinafine.

Lastly, it is worth noting that α-mangostin has demonstrated antiviral properties, particularly in combating severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and human immunodeficiency virus (HIV). The emergence of the coronavirus disease 2019 (COVID-19) pandemic has prompted researchers to identify suitable antivirals to inhibit the spread of SARS-CoV-2. In silico studies on various phytoconstituents indicate that α-mangostin has significant potential as an anti-SARS-CoV-2 agent by inhibiting the main protease (MPro) and angiotensin-converting enzyme 2 (ACE2).^133–136 Molecular docking results reveal that α-mangostin has binding affinities of −8.58 and −4.27 kcal/mol for each receptor, respectively.^11,47,87 These binding affinities are stronger compared to some off-label drugs like lopinavir, nelfinavir, remdesivir, and chloroquine. Regarding HIV, α-mangostin inhibits HIV reverse transcriptase with a binding affinity of −10.51 kcal/mol,⁸⁸ leading to the virus’s failure to convert its RNA genetic material into DNA within host cells.^137,138

Inflammation

In addition to being a potent antioxidant, α-mangostin is increasingly associated with its notable anti-inflammatory properties. In silico studies suggest that α-mangostin effectively inhibits pro-inflammatory enzymes, such as cyclooxygenase I (COX-I) and cyclooxygenase II (COX-II), nuclear factor kappa beta (NF-kB), and AGE-RAGE interaction (Figure 9). In a study conducted by Mohan et al, α-mangostin demonstrated binding affinities of −3.81 and −8.26 kcal/mol towards COX-I and COX-II, respectively.¹⁰¹ The binding affinity of α-mangostin towards COX-I, above −5.50 kcal/mol, indicates a weak binding affinity, implying that α-mangostin likely provides selective inhibition of COX-II only. This finding addresses the issue of non-selective non-steroidal anti-inflammatory drugs (NSAIDs) which often exhibit increased side effects.^139,140

Figure 9 Mechanism of action of α-mangostin as anti-inflammatory predicted using computational studies.

Furthermore, α-mangostin is predicted to have anti-inflammatory properties by inhibiting cell signalling within the immune system. α-mangostin interacts with the p50 subunit of nuclear factor kappa beta (NF-kB) with a binding affinity of −9.64 kcal/mol.¹⁰¹ Deactivation of NF-kB leads to the downregulation of immune system-related cytokine transcription and macrophage stimulation.¹⁴¹ In other findings, a study conducted by Faisal et al also discovered that α-mangostin can intervene in the interaction between advanced glycation end-products (AGE) and their receptor, the advanced glycation end-products receptor (RAGE).^102,142 RAGE is a member of the immunoglobulin superfamily, encoded in the Class III region of the major histocompatibility complex (MHC Class III).¹⁴³ Inhibiting the AGE-RAGE interaction can prevent the transmission of inflammatory signals through the MHC Class III pathway.^142,143 α-mangostin competes with AGE for binding at the active site of RAGE, resulting in a binding affinity of −260.42 kcal/mol.¹⁰²

Mosquito Larvicidal

In addition to its direct application for human treatment, α-mangostin also exhibits significant potential as a mosquito larvicide. Particularly, in the case of dangerous mosquitoes like Aedes aegypti, carrier proteins responsible for cholesterol become essential factors in the process of larval development into adult mosquitoes.¹⁴⁴ Among sterol carrier proteins (SCPs), the type SCP-2 in Aedes aegypti mosquitoes (AeSCP-2) plays the most significant role in mosquito cholesterol metabolism.¹⁴⁵ Therefore, targeting inhibition in AeSCP-2 presents a viable option for finding larvicidal solutions for Aedes aegypti mosquitoes. A study by Kumar et al demonstrates that α-mangostin can compete with cholesterol to bind with AeSCP-2.¹⁰³ The bond formed is robust, with a binding affinity of −13.65 kcal/mol. This value is significantly higher than that of other larvicidal candidates like panthenol, which exhibits a binding affinity of −9.23 kcal/mol for AeSCP-2.

Pain Management

Surprisingly, α-mangostin is known to function as an analgesic. This is indicated in studies by Sani et al and Cui et al, showing that α-mangostin effectively alleviates pain and inflammatory symptoms in rodent models induced by a hot plate and formalin injection.^146,147 While α-mangostin is recognized for its anti-inflammatory properties that contribute to pain relief, the exact mechanism behind α-mangostin’s analgesic effects remains uncertain. Through computational studies conducted by Kim et al, it is predicted that α-mangostin imparts its analgesic effects by activating tandem pore domains in a weak inward-rectifying K⁺ channel (TWIK)-related K⁺ channels (TREK), specifically TREK-1 and TREK-2, and TWIK-related arachidonic acid-activated K⁺ channels (TRAAK). This action inhibits transient receptor potential vanilloid 1 (TRPV1) and reduces tetrodotoxin (TTX)-sensitive Na_V (Na_V1.7), as visualized in Figure 10.¹⁰⁵

Figure 10 Role of α-mangostin in pain management based on computational study.

TRPV1 plays a vital role as a membrane channel in T cells, regulating the influx of calcium ions. In addition to body temperature regulation, TRPV1 also senses scalding heat and pain (nociception). Inhibiting TRPV1 suggests the suppression of receptor potentials.^148,149 TREK-1, TREK-2, and TRAAK are key K⁺ channels at the nodes of Ranvier for rapid action potential conduction in mammalian myelinated afferent nerves.^150–152 On the other hand, Na_V1.7 belongs to the family of voltage-gated sodium channels, facilitating sodium influx under physiological conditions when the nerve membrane is depolarized.^153,154 The activation of TREK/TRAAK and the inhibition of Na_V1.7 indicate membrane stabilization and the suppression of action potential firing in dorsal root ganglion (DRG) neurons.

Rheumatoid Arthritis

In the context of rheumatoid arthritis (RA), the crucial role of immune system regulation is paramount. One of the factors that can alleviate hyperimmunity in RA is peroxisome proliferator-activated receptor γ (PPAR-γ).¹⁵⁵ The activation of PPAR-γ is known to actively contribute to the overexpression of sirtuin 1 (SIRT-1) through an unknown cross-talk mechanism. This overexpressed SIRT-1 subsequently leads to the final suppression of immune system activity by preventing macrophage/monocyte polarization.¹⁵⁶ Interestingly, the pharmacological targeting potential of PPAR-γ in RA is substantial, given that upregulation of PPAR-γ is observed in RA conditions.¹⁵⁷ α-mangostin is known to interact strongly with PPAR-γ, exhibiting a binding affinity of −6.7 kcal/mol.¹⁰⁶ This finding is reinforced by in vivo testing conducted by Wu et al, showing that the group administered with α-mangostin demonstrates overexpression of SIRT-1, aligning with the expected mechanism for RA inflammation management.

Future Perspective

The utilization of computational studies continues to evolve over time, offering extensive opportunities for investigating the therapeutic potential of α-mangostin. This review highlights the software tools employed to predict α-mangostin’s efficacy, including ArgusLab, AutoDock, AutoDock Vina, Glide, HEX, and MOE. However, there is still a plethora of software yet to be maximized for evaluating α-mangostin’s mechanisms in specific clinical pathologies. Offline software such as GOLD,¹⁵⁸ FlexX,¹⁵⁹ rDOCK,¹⁶⁰ and UCSF Dock,¹⁶¹ along with web server-based software like PatchDock,¹⁶² HADDOCK,¹⁶³ ClusPro,¹⁶⁴ and FireDOCK,¹⁶⁵ provide promising avenues for exploring the best-suited software in terms of validity and compatibility for α-mangostin’s applications. In its development, the correlation between computational study results and wet lab studies has also continually improved. One significant advancement is in the enhancement of QSAR (Quantitative Structure-Activity Relationship) validation, a method that constructs predictive models of molecular structures based on activity in wet lab studies. Initially, QSAR relied solely on molecular descriptors (2D QSAR); however, it has now transitioned to 3D QSAR, which also incorporates spatial factors that influence molecular interactions.¹⁶⁶

Furthermore, beyond the vast possibilities in software exploration, investigating α-mangostin’s efficacy for various other pathological conditions is imperative. This is underscored by the high likelihood of in vitro and in vivo findings demonstrating the exceptional antioxidant and anti-inflammatory properties of α-mangostin. As known, hyperoxidative and inflammatory conditions can correlate with secondary diseases such as cancer and other metabolic disorders.^167–170 Hence, the continued utilization of computational studies for α-mangostin’s role in pathologies related to oxidative stress and inflammation remains highly necessary. Ultimately, it is expected that a comprehensive database regarding the therapeutic efficacy and mechanisms of α-mangostin will be established based on the overall pharmacological activity predictions from computational studies.

Similarly to the success of computational studies in identifying more effective derivatives of artemisinin, various computational research efforts have also been directed towards finding more effective derivatives of α-mangostin with improved pharmacokinetic profiles.¹⁷¹ For instance, studies conducted by Maulana et al and Mardianingrum et al have sought to discover α-mangostin derivatives with enhanced efficacy in anti-diabetic and anti-breast cancer applications.^83,94 Thus, a potential strategy for drug development from natural sources encompasses not only the discovery of existing compounds but also the search for more optimal derivatives.

Conclusion

α-Mangostin exhibits a wide range of therapeutic potentials, as evidenced by computational studies. The predicted benefits include antioxidant, anti-inflammatory, Alzheimer’s disease treatment, antidiabetic, antimicrobial agent, anticancer, chronic periodontitis management, pain relief, chronic kidney disease treatment, and rheumatoid arthritis. Additionally, intriguing findings suggest that α-mangostin functions as a larvicide against Aedes aegypti mosquito larvae. Each pharmacological activity of α-mangostin is associated with specific target mechanisms, validated by its binding affinity to target receptors. However, it is important to emphasize that this review does not intend to present α-mangostin as a “panacea” or a universal cure for all health conditions. Rather, it aims to guide researchers towards further investigations to substantiate the clinical applications of α-mangostin. It is important to recognize that in silico studies are a crucial component of the early stages in drug design and development. Further clinical research is necessary to elucidate the effectiveness of this compound before it can be translated into clinical practice.